The consistency among lithium-ion battery cells is critical to a battery pack. Due to the difference in characteristics like voltage, internal resistance, capacity, etc. among lithium-ion battery cells of the same model and specification, when a battery pack including multiple battery cells is utilized in an electric vehicle, it would not be able to hold the original level of performance of each single cell. As such, it will seriously affect the application of the battery pack in electric vehicles. Therefore, the consistency of characteristics among battery cells becomes a critical factor in the lifecycle of a battery pack, and lithium-ion battery grading is one of the most important processes in automatic manufacture of battery cells. Lithium-ion battery grading is to categorize batteries in accordance with certain criteria such that batteries of similar performance are grouped together.
Most existing techniques use a battery charging/discharging curve to grade lithium-ion batteries. However, it is hard to accurately grade batteries by such a method. Some techniques propose a grading method based on the length of a horizontal segment of a battery charging/discharging curve. However, the length of a horizontal segment may not be able to thoroughly evaluate the characteristics of a battery. In other techniques, a discretely measurement curve of a battery is first fitted with a polynomial, and then the battery is graded by comparing and filtering the coefficients of the polynomial. However, precise curve fitting requires a high order polynomial, and hence the fitting requires very complicated calculations. Moreover, when grading by polynomial coefficients, any nuance in a high order coefficient may cause a large difference in the curve. As a result, it is hard to control the grading precision.
In view of the above, existing techniques still need improvement and further development.